ZipReader:filenameEncoding:function

[nenge123]

the config can be set a function? replace decodeText
filenameEncoding:string replace to new TextDecoder(str).decode => filenameEncoding(uint8,encoding)

filenameEncoding:function = filenameEncoding(uint8,encoding)

[nenge123]

or export method
open zip.decodeText, we can edit the decodeText to auto decode function.

because zip.js have more cdn link now

[gildas-lormeau]

Thank you for the suggestion. I agree that these options could be useful. It's implemented in the version 2.7.42.

[nenge123]

decode(rawFilename, filenameUTF8 ? CHARSET_UTF8 : filenameEncoding || CHARSET_CP437)

=>

decode(rawFilename, filenameUTF8 ? CHARSET_UTF8 : filenameEncoding || CHARSET_CP437) || decodeText(rawFilename,CHARSET_CP437)

if options.decodeText nothing return ,will no file return
or some one need the 'cp437' decode;

getEntries({
    filenameEncoding:'utf-8',
    decodeText(buf,encoding){
        let text = new TextDecoder().decode(buf);
        let newbuf = new TextEncoder().encode(text);
        console.log(encoding,buf,text,newbuf.byteLength);
        return newbuf.byteLength>buf.byteLength ? new TextDecoder('gbk').decode(buf):text;
    }
});
//file is not utf8
//console.log(encoding,buf,newbuf.byteLength);
//utf-8 ,Uint8Array(37),69
//cp437 ,Uint8Array [buffer: ArrayBuffer(366)],0

[gildas-lormeau]

I'm sorry but I can't understand the problem. In what case could the string be empty? Do you have a specific example of a file name that would cause this problem?

[nenge123]

@gildas-lormeau
when user decodeText have new TextDecoder(encoding).decode(buf)

getEntries({
    filenameEncoding:'utf-8',
    decodeText(buf,encoding){
        return new TextDecoder(encoding).decode(buf);
    }
});

[gildas-lormeau]

But buf is not empty, is it?

[nenge123]

you can try any file.
encoding !== utf-8 the comment is cp437

[nenge123]

@gildas-lormeau

new TextDecoder('cp437').decode(new Uint8Array([12]))
//Uncaught RangeError: Failed to construct 'TextDecoder': The encoding label provided ('cp437') is invalid.
    at <anonymous>:1:1

[gildas-lormeau]

You have to use this code (or another third-party library) to decode strings encoded in CP437: https://github.com/gildas-lormeau/zip.js/blob/master/lib/core/util/cp437-decode.js. See https://developer.mozilla.org/en-US/docs/Web/API/TextDecoder/encoding

[gildas-lormeau]

Having written the previous answer and after some thought, I can understand that dealing with CP437 encoding on your end is a pain. I've implemented your proposal. If the function encodeText or decodeText returns undefined then the corresponding default function in zip.js will be called as a fallback function.

I should publish a new version in the coming days. Feel free to test the implementation on the master branch meanwhile.